Oliver Trojak, University of Southampton
Nina Vaidya, University of Southampton
Koosha Nassiri NAzif, Stanford University
Nina Vaidya, Lecturer, University of Southampton
Compared to traditional cells, ultra-thin cells of ~1um thickness have power/weight ratios several orders of magnitude higher: enabling most efficient use of launch system constraints. Operation in space requires shielding from damaging radiation present in the space environment. Conventional cells achieve this by using thick and bulky cover glasses, which decreases their specific power further.
Devices using transition-metal dichalcogenides (TMDs) as the energy material can take advantage of the higher absorption coefficient to use thinner layers of active material, reducing overall cell thickness. At these reduced thicknesses, high energy radiation is able to pass through without absorption, and minimal energy deposition; however particles at the extremely low energy end of the radiation spectrum present a hazard: they are able to penetrate through the passivation layers, dump energy and embed in the active layer, and cause damage. As traditional cells are much thicker and used a thick cover glass, the low energy space environment was of no concern as it was automatically blocked, however as devices push into the micron and nanometre regime where high energies are able to pass through the cell, the make-up of this low energy portion of the spectrum becomes increasingly important.
Previous reports with perovskites show that electron radiation causes substantially less damage than proton radiation, which was corroborated by scoping simulations in CASINO with our device structure, leading to a focus on proton radiation tests. Incident proton radiation was simulated using the SRIM software package. Cells with only a thin (~70nm) passivation coating were vulnerable to ~22keV protons, an area where there is relatively poor test data and protocols for testing, and where the flux is many orders of magnitude higher than at 100keV (where data and protocols are present). The incident energy that the cells are vulnerable to can be tuned by adding a thin shielding layer of SiO2 in front of the cell of less than 500nm.
The shielding layer is multi-functional. Due to the differences in refractive index between the WSe2 active layer, the MoOx passivation layer, and the SiO2 shielding layer, the absorption in the WSe2 layer is sensitive to the thicknesses of layers in the stack – allowing absorption to be optimized by varying the shielding thickness. By adding a shielding layer of 540nm of SiO2 on top of 50nm of MoOx passivation, the proton vulnerability energy of the cell was shifted to 100keV, representing several orders of magnitude radiation flux decrease. The absorption of the active layer, meanwhile, increased from ~0.6 to ~0.8.
After exposure to 88keV protons at 1012cm-1 fluence, the efficiency of unshielded was observed to decrease by 12%. This small degradation is reversible: thermal annealing at 90°C was able to almost fully recover the cell (1% degradation). This temperature is achievable during standard cell operation, allowing continuous or periodic repair of accumulated radiation damage
We report on space radiation tested ultra-thin WSe2-based solar cells for space use, where the stack has been designed for multifunctionality of the layers, to minimize dead-mass within the device, and maximize its specific power. Annealing at operational temperatures is shown to recover most of the lost device performance after space radiation exposure, suggesting in-situ repair of the devices is possible during operation.